111–120 of 154 posts
Does the Vision API call back to apple servers in any fashion? Like how google on-device voice recognition APIs will call back to Google when you are online (unless you explicitly pass flags to force it in offline mode). If so, is there any risk in getting your account suspended or ip range banned somehow because of this, for example?
Love the hackiness of this - however, the vision framework is available on Desktop macs as well - https://developer.apple.com/documentation/vision and specifically: https://developer.apple.com/documentation/vision/vnrecognize...
> My preliminary speed tests were fairly slow on my Macbook. However, once I deployed the app to an actual iPhone the speed of OCR was extremely promising (possibly due to the Vision framework using the GPU). I was then able to perform extremely accurate OCR on thousands of images in no time at all, even on the budget iPhone models like the 2nd gen SE. He does mention running it on a macbook
Earlier quoted context omitted.
Author here: KnowYourMeme is one of many sites that memes are continually ingested from (any site that has memes I try to ingest regularly) :)
Amazing work! Also, thank you for making that feed on the main page, been laughing for a while here :D
Bonus: Index from a lot of sources to help track a meme's origin.
This type of thing is on my long list of "can somebody else please do this already".
This title really undersells the absolute insanity of the described solution. This is a beautiful example of "if it's stupid, but it works, it's not stupid." The justification is very convincing. One thing I'm curious about: how did you build your corpus of meme images and videos?
(Not the author) Maybe they leveraged https://knowyourmeme.com/ ? But that can't possibly have all the random memes, could it?
I have a "hackish but works for me" meme database: I use my Telegram "self chat" to send memes I like to myself, and I tag them with the kind of words I'm likely to search for when looking for them later. Works great for me. It's kind of like trying to come up with a good Google search phrase, based on how other people must have phrased something, but relying on knowledge of how you phrase things instead.
Out of curiosity, how does your Image Similarity Search works ? Are you also using some feature of Apple's Vision framework, or running some ML model on your linode instance ?
Short TL;DR: It runs off my home server running a large vector database (opendistro): https://opendistro.github.io/for-elasticsearch-docs/docs/knn...
Absolutely amazing on the tech side!!! Now, after reading the article, I gave your search engine a try. I was looking for that futurama its a trap meme (pretty much pops up on any image search here https://www.google.com/search?q=futurama+its+a+trap ) The problem is, the search engine you built is now very text-heavy, which seems to be usually very unconnected to the actual meme. So, searching for "its a trap" did no…
I'd also like to figure out how to turn an image into a description of whats in it. My ML/tensorflow knowledge is very weak though, so I still have a lot to learn here.